Fluid Mechanics Lecture Notes

Fluid Mechanics and Hydraulics
Prepared by : Dr.Ravikumar Ganti, ravikumar.ganti@aiim.ac.in
Fluid Mechanics Lecture Notes
Module 3- Fluid Kinematics:
Source: 1. http://nptel.ac.in/courses/105103095/25
2. http://www.msubbu.in/ln/fm/
Fluid flow methods of analysis of fluid motion, Streamlines, Path lines, Streak lines
and Stream tubes. Types of fluid flow-Steady and unsteady flow, Uniform and non-
uniform flow, Laminar and turbulent flow, Reynolds number, Reynolds experiment,
Rotational and Irrotational flow, Subcritical, critical and Supercritical flow,
Compressible and Incompressible flow, One, Two and three dimensional flow,
Circulation and vorticity, Stream function and Flownet.
Lecture 1: Fluid flow methods of analysis of fluid motion

Introduction
The fluid kinematics deals with description of the motion of the fluids without
reference to the force causing the motion.
Kinematics is defined as that branch of science which deals with motion of particles
without considering the forces causing the motion. The velocity at any point
in a flow field at any time is studied in this branch of fluid mechanics. Once
the velocity is known, then the pressure distribution and hence forces acting
on the fluid can be determined. In this chapter, the methods of determining
velocity and acceleration are discussed.
Thus it is emphasized to know how fluid flows and how to describe fluid motion.
This concept helps us to simplify the complex nature of a real fluid flow.
When a fluid is in motion, individual particles in the fluid move at different

velocities. Moreover at different instants fluid particles change their positions. In
order to analyse the flow behaviour, a function of space and time, we follow one of
the following approaches
1. Lagarangian approach
2. Eularian approach
the Lagarangian approach a fluid particle of fixed mass is selected. We follow the
fluid particle during the course of motion with time (fig. L-16.1)
Fig. L-16.1
The fluid particles may change their shape, size and state as they move.
As mass of fluid particles remains constant throughout the motion,
the basic laws of mechanics can be applied to them at all times.
The task of following large number of fluid particles
Is quite difficult.
Therefore this approach is limited to some special applications for example re-
entry of a spaceship into the earth's atmosphere and flow measurement system
based on particle imagery.
In the Eularian method a finite region through which fluid flows in and out is used.
Here we do not keep track position and velocity of fluid particles of definite mass. But,
within the region, the field variables which are continuous functions of space
dimensions ( x , y , z ) and time ( t ), are defined to describe the flow. These field
variables may be scalar field variables, vector field variables and tensor quantities. For
example, pressure is one of the scalar fields. Sometimes this finite region is referred as
control volume or flow domain.
For example the pressure field 'P' is a scalar field variable and defined as
Velocity field, a vector field, is defined as
Similarly shear stress is a tensor field variable and defined as

Note that we have defined the fluid flow as a three dimensional flow in a Cartesian co-
ordinates system.
Steady and
unsteady
flows
The flow in which the field variables don't vary with time is said to be steady flow.
For steady flow,
Or
It means that the field variables are independent of time. This assumption simplifies
the fluid problem to a great extent. Generally, many engineering flow devices and
systems are designed to operate them during a peak steady flow condition.
If the field variables in a fluid region vary with time the flow is said to be unsteady
flow.
One, two and three dimensional flows
Although fluid flow generally occurs in three dimensions in which the velocity field vary
with three space co-ordinates and time. But, in some problem we may use one or two space
components to describe the velocity field. For example consider a steady flow through a long
straight pipe of constant cross-section. The velocity distributions shown in figure are
independent of co-ordinate x and and a function of r only. Thus the flow field is one
dimensional.
Fig. L-16.3
But in the case of flow over a weir of constant cross-section (), we can use two co-ordinate
system x and z in defining the velocity field. So, this flow is a case of two dimensional flow.
The reduction of independent space variable in a fluid flow problem makes it simpler to
solve.
Laminar and
Turbulent flow
In fluid flows, there are two distinct fluid behaviours experimentally observed.
These behaviours were first observed by Sir Osborne Reynolds. He carried out a
simple experiment in which water was discharged through a small glass tube
from a large tank (the schematic of the experiment shown in Fig.). A colour dye
was injected at the entrance of the tube and the rate of flow could be regulated
by a valve at the out let.
When the water flowed at low velocity, it was found that the die moved in a
straight line. This clearly showed that the particles of water moved in parallel
lines. This type of flow is called laminar flow, in which the particles of fluid moves
along smooth paths in layers. There is no exchange of momentum from fluid
particles of one layer to the fluid particles of another layer.
This type of flow mainly occurs in high viscous fluid flows at low velocity, for
example, oil flows at low velocity. Fig. shows the steady velocity profile for a
typical laminar flow.
When the water flowed at high velocity, it was found that the
dye colour was diffused over the whole cross section. This
could be interpreted that the particles of fluid moved in very
irregular paths, causing an exchange of momentum from one
fluid particle to another. This type of flow is known as turbulent
flow. The time variation of velocity at a point for the turbulent
flow is shown in Fig.
Fig-L16.5 : Velocity profile for laminar flow L-16.6

profil
flow
It means that the flow is characterized by continuous random

fluctuations in the magnitude and the direction of velocity of the fluid
particles.
Example 1 :
A velocity field is defined by u = 2 y2, v = 3x, w = 0. At point (1,2,0), compute the a) velocity, b) local
acceleration and a) convective acceleration.
Solution :
Given velocity field, u = 2y 2 ; v = 3x; w = 0 so,
a) Thus,
b) Now from the above equation we can observe that
Which implies the local acceleration is zero.
c) Also from the above equation we have the acceleration component as follow
Types of Fluid Flow
Uniform and Non-uniform flow : If the velocity at given instant is the same in both
magnitude and direction throughout the flow domain, the flow is described as uniform.
Fig. L-16.2 : Uniform

and Non-uniform flow.
Mathematically the velocity field is defined as , independent to space dimensions ( x , y , z ).
When the velocity changes from point to point it is said to be non-uniform flow. Fig.() shows
uniform flow in test section of a well designed wind tunnel and ( ) describing
non uniform velocity region at the entrance.

Velocity Field Print this page
Introduction
Consider a uniform stream flow passing through a solid cylinder (Fig.). The typical velocities at different
locations within the fluid domain vary from position to position at a particular time t . At different time
instants this velocity distribution may change. Keeping this observation in mind, the velocity within a flow
domain can be represented as function of position ( x , y , z ) and time t .
Fig. L-17.1
In the Cartesian co-ordinates the variation of velocity can be represented as a vector whe
u , v , w are the velocity scalar components in x , y and z directions respectively.
The scalar components u , v and w are dependent functions of position and time. Mathematically we can
express them as
This type of continuous function distribution with position and time for velocity is known as velocity field.
is based on the Eularian description of the flow. We also can represent the Lagrangian description of
velocity field.
Let a fluid particle exactly positioned at point A moving to another point during time interval . The
velocity of the fluid particle is the same as the local velocity at that point as obtained from the Eulerian
description
At time t , particle at x , y , z
At time , particle at
This means that instead of describing the motion of the fluid flow using the Lagrangian description, the use
of Eularian description makes the fluid flow problems quite easier to solve. Besides this difficult, the
complete description of a fluid flow using the Lagrangian description requires to keep track over a large
number of fluid particles and their movements with time. Thus, more computation is required in the
Lagrangian description.
The
Acceleration
field
At given position A, the acceleration of a fluid particle is the time derivative of the
particle's velocity.
Acceleration of a fluid particle:
Since the particle velocity is a function of four independent variables ( x , y , z and

t ), we can express the particle velocity in terms of the position of the particle as
given below
Applying chain rule, we get
Where and d are the partial derivative operator and total derivative operator
respectively.
Velocity Field Print thi
The time rate of change of the particle in the x -direction equals to the x -component of
velocity vector, u . Therefore
Similarly,
As discussed earlier the position vector of the fluid particle ( x particle , y particle , z
particle ) in the Lagranian description is the same as the position vector ( x , y , z ) in the
Eulerian frame at time t and the acceleration of the fluid particle, which occupied the
position ( x , y , z ) is equal to in the Eularian description.
Therefore, the acceleration is defined by
Vector form
where is the gradient operator.
The first term of the right hand side of equation represents the time rate of change of
velocity field at the position of the fluid particle at time t . This acceleration
component is also independent to the change of the particle position and is referred
as the local acceleration. However the term accounts for the affect of the
change of the velocity at various positions in this field. This rate of change of velocity
because of changing position in the field is called the convective acceleration.
The
Acceleration
field
At given position A, the acceleration of a fluid particle is the time

derivative of the particle's velocity.
Since the particle velocity is a function of four independent variables ( x , y , z

and t ), we can express the particle velocity in terms of the position of the
particle as given below
Where and d are the partial derivative operator and total derivative
operator respectively.
The Acceleration field
At given position A, the acceleration of a fluid particle is the time derivative of
the particle's velocity.
Since the particle velocity is a function of four independent variables ( x , y , z

and t ), we can express the particle velocity in terms of the position of the
particle as given below

Where and d are the partial derivative operator and total derivative operator
respectively.
The time rate of change of the particle in the x -direction equals to the x -component
of velocity vector, u . Therefore
Similarly,
As discussed earlier the position vector of the fluid particle ( x particle , y particle , z
particle ) in the Lagranian description is the same as the position vector ( x , y , z ) in
the Eulerian frame at time t and the acceleration of the fluid particle, which occupied
the position ( x , y , z ) is equal to in the Eularian description.
Therefore, the acceleration is defined by
But in the case of the particle B , it can easily be seen that the particle rotates in
clockwise direction near the obstruction. This results due to the presence of the
velocity gradient at that region. So, this type of motion of a fluid particle is known as
rotation.
The particle C moves in the region of high velocity gradient. Therefore, the particle is
deformed volumetrically and is also undergoes angular deformation because of non-
uniform distribution of velocity in the path x and y directions.
In short the types of primary motion of a fluid particle are described in four ways: (a)
translation (b) rotation (c) linear deformation and (d) angular deformation.
Rotatio
n
Consider a two dimensional fluid particle motion in a fluid flow domain. The flow velocity
at point A ' of the particle is expressed as
As per the continuum hypothesis the velocity components u and v are continuous functions
of space and time. The velocity at point A can be expressed using the Taylor series
Neglecting the second and higher order terms in the above expression we obtain
Similarly the velocity components at point B and can be derived.
Deformation of fluid particles
Fig. L-18.3: Rotation of fluid particle

The pure rotation of the element is resulted from v -velocity component at point and the u -veloc
at point B .
The angle rotated during time
Deformation of fluid particles
Print this page
similarly,
The negative sign has been introduced because of clockwise rotation.
The average rotation angle is
Thus, the rate of rotation in the x and y planes becomes
In three dimension we can express rate of rotation or angular velocity in vector form as
Linear
deformation
In fluid mechanics the rate of linear deformation is emphasized instead of linear

deformation in solid mechanics. The rate of linear deformation or linear strain rate is
the rate of increased or decreased length per unit length.
Consider two points P and Q located on a fluid particle in the x -direction. The velocity
at pint P and Q at time t are u and respectively. During time , P moves to

and Q to . The rate of linear deformation is
Thus
Similarly linear strain rate in other directions are

Prev
Angular deformation :
As shown in Fig. L-18.4, angular deformation at point P is defined as the half of the rate of the angle
decreased between two mutually perpendicular axes.
Fig. L-18.4 : Angular deformation of fluid particle

Deformation Print this page

of fluid
particles
The angular deformation in the xy plane is
Note that is in the clockwise direction. Extending to three dimensions the shear strain
rate is given by
Rotation
Consider a two dimensional fluid particle motion in a fluid flow domain.

The flow velocity at point A ' of the particle is expressed as
As per the continuum hypothesis the velocity components u and v are

continuous functions of space and time. The velocity at point A can be
expressed using the Taylor series
Neglecting the second and higher order terms in the above expression
we obtain
Similarly the velocity components at point B and can be derived.

Deformation of fluid particles P

r
i
n
t
t
h
i
s
p
a
g
e
Fig. L-18.3: Rotation of fluid

particle
The pure rotation of the element is resulted from v -velocity component at point
and the u -velocity component at point B .
The angle rotated during time

Deformation of fluid particles Print this page
similarly,
The negative sign has been introduced because of clockwise rotation.
The average rotation angle is
Thus, the rate of rotation in the x and y planes becomes
In three dimension we can express rate of rotation or angular velocity

in vector form as
Deformation of fluid particles Print this page
Linear deformation
In fluid mechanics the rate of linear deformation is emphasized instead of linear deformation in
solid mechanics. The rate of linear deformation or linear strain rate is the rate of increased or
decreased length per unit length.
Consider two points P and Q located on a fluid particle in the x -direction. The velocity at pint P
and Q at time t are u and respectively. During time , P moves to and Q to . The
rate of linear deformation is
Thus
Similarly linear strain rate in other directions are

Angular deformation :
As shown in Fig. L-18.4, angular deformation at point P is defined as the half of the rate of
the angle decreased between two mutually perpendicular axes.
Fig. L-18.4 : Angular

deformation of fluid particle
The angle between these two axes decreases from to , as demonstrated

in FigureL-18.4 . The rate of angle , already derived in section() is
Deformation of fluid particles Print this
The angular deformation in the xy plane is
Note that is in the clockwise direction. Extending to three dimensions the shear
strain rate is given by
The Reynolds Transport Theorem Print this

page
Introduction
The basic equations given in section (), involving the time derivative of extensive
properties (mass, linear momentum, angular momentum, energy) are required to
analyse any fluid problem. In solid mechanics, we often use a system representing a
quantity of mass of fixed identity. The basic equations are therefore directly applied to
determine the time derivatives of extensive properties. However, in fluid mechanics it
is convenient to work with control volume, representing a region in space considered
for study. The basic equations based on system approach can not directly applied to
control volume approach.
Fig. illustrates different types of control volume: fixed control volume, control volume
moving at a constant speed and deforming control volume. In this section, it is aimed
to derive a relationship between the time derivative of system property and the rate of
change of that property within a control volume. This relationship is expressed by the
Reynolds Transport Theorem (RTT) which establishes a link between the system and
control volume approaches.
Before deriving the general form of the RTT, a derivation for one dimensional fixed
control volume is explained in the next section.
FLUID KINEMATICS Print this

page
One- dimensional fixed control volume:
Consider a diverging (expanding) flow field bounded by a stream tube. The chosen control volume
is to be fixed between section a' and section 'b'. Note that both the sections are normal to the
direction of flow. At initial time t , System I exactly coincides with the chosen control volume. This
assumption says that the system and control volume are identical at that time. At time ,
System-I has moved in the flow direction at uniform speed and a part of system II has entered
into the control volume.
Let ' N' represent any properties of the fluid (mass, momentum, energy) and then represent the
amount of ' N'' per unit mass (known as intensive property) in a small proportion of the fluid. The
total amount of N ' in a control volume is expressed as
As the system coincides with the control volume at time ' t ', a relation between the system and the
control volume is
At time ,
Using the definition of derivative, we can write,
or
The influx rate of N through the control surface can be computed as
Finally, Equation ( ) becomes

This equation implies that the time rate of change of any extensive property for a
system is equal to the rate of change of that extensive property inside the control
volume plus the net efflux of the property through the control surface. This is known as
Reynolds Transport Equation which relates the change of a property of a system to the
change of that property for a control volume.
Arbitrary Fixed Control Volume
As similar to the previous derivation, consider a fixed control volume with an arbitrary flow pattern
passing through. At time t the system coincides with the control volume which is fixed relative to
the x , y, and z axes. At time , the system has moved and occupies the region II and III as
shown in Fig.L-19.1. Note that the region II is common to the system at both times t and .
The time rate of change of N ' for the system can be given by
Rearranging the above equation we have
region II occupies the same space of control volume and the first term on the right side of the
above equation becomes . The integral for region III approximates the amount of N ' that
has crossed the control surface ABCD shown in Fig.L-19.1.
Fig. L-19.1
Let an area dA on the control surface where a steady flow velocity v is attained during time interval,
, the interface has moved a distance along a direction which is tangential to streamline at that
point.
The volume of the fluid swept across the area dA is
Using the dot product we can define
So, the integral for the region III, is expressed by substituting . Efflux rate through control surface
ABC is
Similarly, the influx rate through control surface ADC can be expressed
Influx rate:
The negative sign indicates influx rate of N pass through the control surface. The net
efflux rate of N through the whole control surface is
Net efflux rate
Collecting the terms of equation ( ) gives the compact from of RTT as
The above equation states that the time rate of change of property N within a system is
equal to the time rate of change of property N within the arbitrary shaped control
volume plus the net rate of efflux of the property across the control surface.
Special
Cases
Control volume moving at constant velocity :
In the case of fixed control volume the velocity field was measured with reference to x , y
, z axes . If the control volume moves at a uniform velocity it is necessary to compute
relative velocity of fluid crossing the control surface. The relative velocity becomes
. The flux terms are to be defined in terms of , but the volume integral
remains unchanged when the control volume does not deform.
The RTT for a uniformly moving control volume is
Fluid
Statics
Exampl
e 1:
Calculate the velocity of water flowing through a pipe system in which the flow first
enters a pipe of 0.2 m diameter and releases to another pipe of 0.1 m diameter. The water
flows at a rate of 2200 litre per hour.
Solution :
Given data: Flow rate Q = 2200 lph
Velocity of water at the first pipe
If A 2 and V 2 are the cross sections and velocity in the second pipe
Using continuity equation,
Substituting the values

Answer:
Fluid
Statics
Example 2:
What will be the change of velocity of water through a nozzle of diameter

0.01m if a change of flow takes place from 1500 lph to 1200 lph?
Solution :
Given data: Flow Rate Q 1 = 1500 lph
Q 2 = 1200 lph
From the continuity equation we get
Substituting the values after the unit conversions of flow rate V 1 =5.3 cm/s
V 2 = 4.2 cm/s
Answer: 5.3 cm/s and 4.2 cm/s
Example
3:
The two scalar components of velocity field are given for two flow system.
Find the third component of the velocity. For case I) u=x 3 +2y 2 +z 3 and v
= -x 2 y yz - xy . And II) u= log (y 2 + z 2 ) and w= log (x 2 +y 2 ).
Solution :
Case I: Velocity field should satisfy the continuity equation. For a three
dimensional flow the equation is
The partial derivatives of u and v are
Substituting these into the continuity equation

Integrating we have,
Answer: -
Case II: As similar to the case I we can derive
Integration of the above equation gives v = f(x,z)

Since u and w are symmetric function of x, y and z , v should also be a
symmetric function.
So,
Reynolds
Transport
Theorem
Example
4:
An incompressible fluid flows past a solid plate. If the x component of the

velocity field is given by u = x 2 y 2 + 2xy , obtain the velocity field and
also the acceleration at point (2, 2).
Solution :
Since the flow is two dimensional, the continuity equation becomes
The partial derivative of u is
Integrating the above equation
Now,
Acceleration:
Acceleration (total) In
which
Then, total acceleration
Computing the dot product and substituting x and y coordinates (2, 2) of the point we get
Acceleration =
Answer:-
Example
5:
The velocity field of an incompressible flow is given by V = 3y 2 i - 8x j . Verify whether the

fluid is in motion. If so, determine the expression for stream function and the direction of the
stream line through (1, 2).
Solution :
From the given data:
Answer:
Example 6:
Derive the equation of stream function and velocity potential for a uniform flow of
stream of velocity 5 m/s at a angle of 30 0 to the x-axis in a two dimensional field.
Solution :
From the given data
Now, from the definition of the stream function we get

Integrating we have
So, the equation
Also,
Integration of the above gives f = -4.33 x +f(y) and f = 2.5 y + f(x)
T hen, f = 2.5y - 4.33x
Exam
ple 7:
The x- component of the velocity field for a two dimensional fluid flow over a plate is
given by u = 3y y 2 .Where y is measured from the surface in a perpendicular direction.
Verify if the flow is rotational. If so find the rotational velocity at (3, 2).
Solution :
Given data : and the flow is two dimensional.
From the continuity equation
Now imposing the boundary condition, at y = 0 v=0 => f(x) = 0.
The rotational component in the z direction
Answer:
Exam
ple 9:
For a two dimensional flow stream function is given as 5(y 2 -x 3 ). Find out the velocity
components and the discharge passing between the streamlines through the point (3,2)
and (4,2)
Solution :
Given data: So, velocity components
At (3,2) and Q = ? 2 ? 1
Answer:-
Exam
ple
10:
The velocity field of a flow is defined as

d.
a. Is the flow steady or unsteady?
b. Can you approximate the flow as a 2-dimensional flow?
c. Determine the acceleration field?
Solution :
Given velocity field
a) As it is clear from the above that the velocity field is a function of time and hence
the flow is unsteady
b) Since the velocity component is not zero in z-direction ( w ) it can't be

approximated to a two dimensional flow.
c) The acceleration component is determined as follow.

http://www.msubbu.in/ln/fm/Unit-II/ContinuityEqun.htm
Continuity Equation
Home -> Lecture Notes -> Fluid Mechanics -> Unit-II
Let us make the mass balance for a fluid element as shown below: (an open-faced
cube)
Mass balance:
Accumulation rate of mass in the system = all mass flow rates in - all mass flow rates
out --> 1
The mass in the system at any instant is x y z. The flow into the system through
face 1 is
and the flow out of the system through face 2 is

Similarly for the fcaes 3, 4, 5, and 6 are written as follows:
Substituting these quantities in equn.1, we get
Dividing the above equation by x y z:
Now we let x, y, and z each approach zero simulaneously, so that the cube shrinks
to a point. Taking the limit of the three ratios on the right-hand side of this equation,
we get the partial derivatives.
This is the continuity equation for every point in a fluid flow whether steady or
unsteady , compressible or incompressible.
For steady, incompressible flow, the density is constant and the equation simplifies
to
For two dimensional incompressible flow this will simplify still further to

Fluid Mechanics Lecture Notes - Module3

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Fluid Mechanics Lecture Notes - Module3

Uploaded by

Copyright:

Available Formats

Fluid Mechanics and Hydraulics

Prepared by : Dr.Ravikumar Ganti, ravikumar.ganti@aiim.ac.in

Module 3- Fluid Kinematics:

Lecture 1: Fluid flow methods of analysis of fluid motion

When a fluid is in motion, individual particles in the fluid move at different

As mass of fluid particles remains constant throughout the motion,

the basic laws of mechanics can be applied to them at all times.

The task of following large number of fluid particles

Velocity field, a vector field, is defined as

Similarly shear stress is a tensor field variable and defined as

One, two and three dimensional flows

Fig-L16.5 : Velocity profile for laminar flow L-16.6

It means that the flow is characterized by continuous random

Given velocity field, u = 2y 2 ; v = 3x; w = 0 so,

b) Now from the above equation we can observe that

Which implies the local acceleration is zero.

Types of Fluid Flow

Fig. L-16.2 : Uniform

Mathematically the velocity field is defined as , independent to space dimensions ( x , y , z ).

non uniform velocity region at the entrance.

Acceleration of a fluid particle:

Since the particle velocity is a function of four independent variables ( x , y , z and

Applying chain rule, we get

Velocity Field Print thi

Therefore, the acceleration is defined by

where is the gradient operator.

At given position A, the acceleration of a fluid particle is the time

Acceleration of a fluid particle:

Since the particle velocity is a function of four independent variables ( x , y , z

Applying chain rule, we get

Acceleration of a fluid particle:

Since the particle velocity is a function of four independent variables ( x , y , z

Applying chain rule, we get

Therefore, the acceleration is defined by

Similarly the velocity components at point B and can be derived.

Deformation of fluid particles

Fig. L-18.3: Rotation of fluid particle

The angle rotated during time

Deformation of fluid particles

Print this page

The negative sign has been introduced because of clockwise rotation.

The average rotation angle is

Thus, the rate of rotation in the x and y planes becomes

In fluid mechanics the rate of linear deformation is emphasized instead of linear

at pint P and Q at time t are u and respectively. During time , P moves to

Similarly linear strain rate in other directions are

Fig. L-18.4 : Angular deformation of fluid particle

Deformation Print this page

The angular deformation in the xy plane is

Consider a two dimensional fluid particle motion in a fluid flow domain.

As per the continuum hypothesis the velocity components u and v are

Similarly the velocity components at point B and can be derived.

Deformation of fluid particles P

Fig. L-18.3: Rotation of fluid

The angle rotated during time

Deformation of fluid particles Print this page

The negative sign has been introduced because of clockwise rotation.

The average rotation angle is

Thus, the rate of rotation in the x and y planes becomes

In three dimension we can express rate of rotation or angular velocity

Deformation of fluid particles Print this page

Similarly linear strain rate in other directions are